How UK banks are looking to use AI and machine learning: RBS, HSBC, Lloyds and Barclays embrace AI

One of the core uses for machine learning in the banking world has been to combat fraud and improve compliance. The technology is ideally suited to the problem as machine learning algorithms can comb through huge transactional data sets to spot unusual behaviour.

"When you know about [fraud] now, something can be done about it," Andrew McCall, chief engineer for big data at Lloyds Banking Group said earlier this year. "If you know about something that happened yesterday, it is not as effective as an anti-fraud mechanism."

Read next: How Stripe plans to combat online payments fraud using machine learning and analytics with Radar

Douglas Flint, chairman of HSBC said at the inaugural International Fintech Conference in April: "Using AI and machine learning to police the financial system is creating opportunities to do things better, to protect customers and ourselves."

Banks have used __computer algorithms to trade stocks and shares since the 1970s, a practice that was partly responsible for the 1987 'Black Monday' stock market crash. But as AI systems get better the banks are naturally looking towards the technology to get an edge over the competition.

Head of group innovation at Barclays Michael Harte said in April that the most obvious use for AI in banking is "in large algorithmic trading". This means "using vast amounts of high velocity data to outsmart the competition and to provide better instruments and value to customers."

It is worth noting that where Harte said customers, he presumably means Barclays' investment clients, rather than everyday consumer banking customers. For more on electronic high frequency trading and how it helped a handful of banks 'rig' the US equity market Michael Lewis' book Flash Boys is well worth a read.

Being able to track transactions in real-time has historically been an issue for major banks that have a huge amount of legacy IT infrastructure. However, getting data in place to be able to track transactions at low-latency would not only give the banks a better view of their customers, it would also give them the dataset required to apply AI and deep learning to provide personalised, value added products to customers as it learns about spending habits over time.

Read next: Lloyds Banking Group prepares for open banking by shifting towards realtime data feeds

The data science team at Lloyds has been working on this problem recently and has found a way to track transactions in near to real-time. Andrew McCall, chief engineer for big data at Lloyds Banking Group said that getting to near real-time data processing within the bank "starts to open up lots of possibilities in terms of machine learning and how we can better serve customers and give them better insight into their own finances."

After receiving a $1.9bn (£1.2bn) fine over money laundering in 2012, HSBC admitted that its money laundering controls were not fit for purpose. The bank said earlier this year that it is using Google Cloud machine learning capabilities for anti-money laundering.

The CIO at HSBC Darryl West said the bank is using machine learning to run "analytics over this huge dataset with great compute capability to identify patterns in the data to bring out what looks like nefarious activity within our customer base. The patterns that we identify are then escalated to the agencies and we work with them to track down the bad guys."

Read next: HSBC turns to Google Cloud for analytics and machine learning capabilties

As startups like the UK-based ComplyAdvantage try to show, AI is ripe for application to tracking money laundering as it is especially good at spotting odd behaviour within large data sets, like banking transactions.

Barclays head of group innovation Michael Harte envisions an AI system that can help the bank create and recommend better banking products to customers on a more personalised basis.

He said that the people within banks responsible for "inventing and maintaining products and services for customers" should be most excited by the technology. "So that what you bring customers is what they need. So tailor specific algorithms to the individual, instead of selling these generic products," he said.

Read next: Meet the UK's new breed of digital mortgage brokers: these UK fintech startups want to fix the 'soul crushing' mortgage process

For example, there is already a growing number of startups that are looking to use machine learning algorithms to help customers find the best mortgage product, such as Trussle and Habito in the UK.

Banks can use machine learning algorithms to analyse an applicant for credit, be that an individual or a business, and make approvals according to a set of pre-defined parameters. These algorithms simply look at a customers credit score, age and postcode to make a decision in seconds.

However there are concerns that algorithms may not be as impartial as people think, and that getting the banks to explain the decision-making of their AI may not be as simple as it seems.

“It’s not entirely clear how to properly equip a watchdog to do the job, simply because we are often talking about very complex systems that are unpredictable, change over time and are difficult to understand, even for the teams developing them,” Brent Mittelstadt of the Oxford Internet Institute told The Guardian.

Royal Bank of Scotland (RBS) has designed a customer service chatbot called Luvo, which is due to launch to customers in 2017 following a trial last year.

Luvo is a natural language processing AI bot which will answer RBS, Natwest and Ulster bank customer's questions and perform simple banking tasks like money transfers. If Luvo is unable to find the answer it will pass a customer over to a member of staff.

Read next: RBS says it's 'twelve months ahead' of other UK banks with Luvo AI customer assistant

The bank said Luvo “talks to you through WhatsApp-type interaction” and what sets it apart from digital assistants like Siri and IKEA’s Ask Anna is its ability to understand context and perform tasks.